1. When the Child Abduction and Custody Act 1985 came into force.

2. See Lowe, and Perry, , “The Operation of the Hague and European Conventions on International Child Abduction between England and Germany Parts I and II” [1998]International Family Law 8 and 51.Google Scholar

3. I.e. our research is confined to applications concerning England and Wales.

4. See e.g. Bruch, , “Child Abduction and the English Courts”, in Bainham, and Pearl, (Eds), Frontiers of Family LawGoogle Scholar, Everall, , “Child Abduction after the Hague Convention” [1990] Fam. Law 169Google Scholar and “The Hague Convention: The Children Act and other Recent Developments” [1992] Fam. Law 164Google Scholar; Schuz, , “The Hague Child Abduction Convention: Family Law and Private International Law” (1995) 44 I.C.L.Q. 771Google Scholar; Silberman, , “The Hague International Child Abduction Convention: A Progress Report” (1994) 57 Law and Contemp. Probs 208Google Scholar, and “Hague Convention on International Child Abduction—A Brief Overview and Case Law Analysis” (1994) 28 Fam. L.Q. 9.Google Scholar

5. In the 1970s Michael W. Agopian studied a sample of 91 families known to the Los Angeles County court system for child snatching (“Parental Child Stealing: Participants and the Victimisation Process” (1980) 2–4 Victimology: An Int J. 5Google Scholar and Parental Child-Stealing(1981)).Google ScholarJanvier, Rosemary F., Kathleen, McCormkk and Rose, Donaldson surveyed 65 parents who had sought help in locating their children from missing children's organisations (“Parental Kidnapping: A Survey of Left-Behind Parents” (1990) Juvenile and Family Court J. 41).Google Scholar Chris Hatcher has carried out longitudinal research into the impact of abduction on parents and children (Hatcher, , Cole, Barton and Loren, Brooks, Families of Missing Children: Psychological Consequences (1992)).Google ScholarGreif, Geoffrey L. and Hegar, Rebecca L. studied a sample of 371 “left-behind” parents who had sought help in searching for their missing children (When Parents Kidnap (1993)).Google Scholar See also Girdner, Linda K. and Hoff, Patricia M. (Eds), Obstacles to the Recovery and Return of Parentally Abducted Children: Final Report (1992)Google Scholar, and Joan, Fisher, Missing Children Research Project: VoL 1 Findings of theStudy (1989).Google Scholar

6. An examination of the Child Abduction Unit's records for 1994, 1995 and 1997 to date showed that there was nothing unusual about the rates of use of the two Conventions in 1996, or about the remedies applied for, or the countries most frequently involved in the applications. Since the coming into force of the Conventions official statistics show that there has been a steady increase in the numbers of applications dealt with each year, from 16 in 1986, to 171 in 1990, 273 in 1993 and 372 in 1996. Note, however, that the number of countries party to each Convention has also increased: in 1986 the Hague Convention was in force between only 6 countries and the European between 7, by the end of 1996 these figures had increased to 45 and 20 respectively.

7. An “incoming” application is an application received by the central authority for England and Wales, referring to a child or children brought into England and Wales.

8. An “outgoing” application is an application made by the central authority for Englandand Wales, referring to a child or children taken from England and Wales.

9. See Appendix 1, Fig.A1. Note that statistics were not provided by every central authority.

10. See Appendix 1, Table A1 and A2 for the 1996 position.

11. Greif and Hegar, op. cit. supra n.5.

12. See Appendix 1, Table A3.

13. Although when European Convention applications are looked at as a proportion of all applications both France and Portugal were higher users than England and Wales in 1996.

14. The equivalent figures for 1994 were 94% Hague and 6% European, for 1995 they were 90% Hague, 10% European, and for 1997 up to 22 July they were 89% Hague and 11% European. Percentages have been rounded up throughout.

15. See e.g. the figures in Appendix 1, Table A3.

16. In fact, an access application under the European Convention may well be more effective since its provisions on access are generally considered superior to those of the Hague Convention.

17. See S v. H (Abduction: Access Rights) [1997] 1 F.L.R. 971.Google Scholar

18. One possible explanation being the level of trust the two countries have in each other's legal systems, while another factor is the Irish central authority's apparent practice of refusing to accept applications concerning access under the Hague Convention (although only 2 of the 6 applications made to Ireland in 1996 under the European Convention were access applications).

19. See Lowe, , “Problems Relating to Access Disputes under the Hague Convention on International Child Abduction” (1994) 8 Int.J. Law and the Family 374.Google Scholar

20. In Re G (A Minor) (Enforcement of Access Abroad) [1993] Fam. 216 (CA).Google Scholar

21. I.e. whereby Convention applicants are neither means nor merits tested: see Civil Legal Aid (General) Regulations 1989, reg.14.

22. Although some of these children were the subject of more than one application. In all there were 28 applications in 1996 concerning the same child or children. A child who was the subject of more than one application has only been counted once. This figure of 559 children compares to the 519 children involved in the Greif and Hegar study, op. cit. supra n.5.

23. In Greif and Hegar's study, ibid, 75% of the children were 6 years old or less. In our sample 56% of the children were aged 6 or under. This difference may in part be accounted for by the fact that their study was of domestic as well as international abductions, while ours involved international abductions only: Greif and Hegar also found that children taken abroad tended to be older.

24. 93 of the single children were involved in “incoming” applications and 119 “inoutgoing” applications, while equal numbers of children were involved in “incoming” and “outgoing” sibling group abductions. For a detailed analysis of applications concerning single children and sibling groups, see infra Section C.

25. Agopian (1980) and (1981), Greif, and Hegar, , all op. cit. supra n.5.Google Scholar

26. Again, excluding multiple applications concerning the same child or children.

27. In the American research, in the samples studied by both Greif and Hegar, and Janvier et al. (both supra n.5) about 45% of the abductions were perpetrated by the child's mother while around 55% were carried out by fathers. In Agopian's sample (supra n.5)around 71% of the abductors were fathers and only about 29% mothers, almost the exact opposite to the situation found among our 1996 sample.

28. Such information being available in just under 41% of the cases.

29. The term “snatched” is used here as a shorthand expression to cover cases where some type of physical violence has been used to achieve the abduction. There were 6 such cases, in 4 of which the abductor was the childl's father, 1 the child's mother and 1 the child's grandparents.

30. The Third Meeting of the Special Commission to Discuss the Operation of the Hague Convention on the Civil Aspects of International Child Abduction held at The Hague in Mar. 1997.

31. For the oblique reference to the discussion see para.59 of the official report, Report of the third Special Commission meeting to review the operation of the Hague Convention on the Civil Aspects of International Child Abduction (17–21 03 1997) (Hague Conferenceon Private International Law, 08 1997).Google Scholar

32. This was a common scenario among Greif and Hegar's sample, op. cit. supra n.5.

33. As to which, see infra Section B.3(b)(ii).

34. The identity of the abductor was not ascertainable in the remaining 7% of the cases. Note that these figures were arrived at after an examination of the CAU's own records for that year, and that the files themselves were not examined. While it is possible to surmise from these records that the abductor was the child's father or mother, we cannot be certain, since only names are given and not relationship. Thus, the figure given for mothers above might well include some grandmothers or aunts and, similarly, the figure for fathers may include other male relatives. However, bearing in mind that most abductions dealt with by the Unit are carried out by parents, the figures are likely to be reasonably accurate.

35. The mother's age was ascertainable in 231 of the 252 cases in which the mother was the abductor.

36. The father's age was known in 81 of the 95 cases where the father was the abductor.

37. In Agopian's research, op. cit. supra n.5, the abductors were mostly aged between 27 and 36, while the average age of abducting parents involved in international abductions in Greif and Hegar's study, op. cit. supra n.5, was 36.

38. Greif and Hegar, ibid.

39. We are aware of the limitations of relying on the abductors' nationality in this way, when it may well be that nationals of a particular country in fact have cultural or family links elsewhere. The link between nationality and “home” becomes even more tenuous in the case of truly peripatetic families.

40. But over a quarter of these (9% of all the parents) had initiated divorce proceedings apparently as a result of the abduction.

41. The marital status of the remaining 4% was not known. These figures compare with those in both Greif and Hegar's and Janvier's samples; they respectively recorded 17% and 9% of “left-behind” parents as still married at the time of the abduction; 43% and 40% as divorced; 27% and 38% as separated; and 13% and 12% respectively as never married to the abductor.

42. The parties' prior living arrangements were known in only 7 of the 54 cases involving abducting mothers who were not married to the child's father; in 3 of these cases the abduction took place during a contact visit, so the parties could not have been living together at the time, and in the remaining 4 it was clear that the parties were cohabiting at the time of the abduction. In the 17 cases where the abductor was an unmarried father, 4 abductions took place during a contact visit, while none were known to have taken place while the parents were living together.

43. These were: two aunts, one paternal, one maternal; one maternal uncle with a residence order, and one nanny. In 2 cases the abductor was not identified on the file.

44. The numbers respectively are 60 to 141.

45. There were 16 abductions of single children over 10 years old by fathers, and 14 by mothers.

46. For the first age group, the proportion of abductions by mothers to abductions by fathers was over 5: 1, for the second age group this had gone down to just over 3: 1, and for the third age group it had decreased to 2: 1.

47. There were 305 boys and 254 girls.

48. See particularly Art.11 of the Hague Convention, which enjoins judicial andadministrative authorities to act expeditiously. The importance of speed is also emphasised in the Explanatory Report (the Elisa Pérez-Vera Report) on the Convention (Hague 1982) at e.g. para. 104.

49. The resulting sample size is 336.

50. The CAU sets itself an 80% target of forwarding “incoming” cases to solicitors within 24 hours, and according to Wall J “it invariably achieves 100% rate”. His Lordship also mentioned that the average-turnaround lime between receipt of an “incoming” application and the final order is 6 weeks and for applications that go before the Court of Appeal the average turnaround is 15 weeks: Re S (Child Abduction: Delay) [1998] 1 F.L.R. 651.Google Scholar The need to maintain the momentum of speedy disposals in the event of an appeal was emphasised by Thorpe, LJ in Re HB (Abduction: Children's Objections) [1998] 1 F.L.R. 422, 427 (CA).Google Scholar Our findings confirm the overall turnaround figure, but are not broken down into cases which are or are not appealed.

51. See Appendix 3, Table A6, which sets out the outcomes reached within time periodsup to 32 weeks.

52. In fact we identified a third category, namely, “other resolutions” which included the application being rejected by the requested central authority, cases where the child was taken to another country and cases which for various reasons petered out. Since we are interested here in the speed of the Convention machinery the length of these cases is of little interest.

53. This group includes judicial returns, judicial refusals to return, registration of custody and access orders, refusal to register orders, the making of access orders, and cases where the Convention proceedings were stayed or dismissed by the court.

54. This category comprises voluntary return of the child and other voluntary agreements not necessarily involving a return.

55. There were 3 completed “incoming” European Convention applications, of which 1 was completed within 3 weeks of the application being made, 1 within 7 weeks, while the third took 35 weeks to reach a conclusion; and 9 completed “outgoing” European Convention cases, lasting between 3 and 38 weeks.

56. Of the “incoming” Hague applications resolved within 2 weeks, 56% were judicially resolved (all by orders for return of the child), 20% resulted in a voluntary return, 20% resulted in the child being taken to another Convention country, and in 1 case an access order was forwarded to the requesting central authority. Among the “outgoing” applications resolved within 2 weeks or less, 12.5% resulted in a judicial return of the child, 58% resulted in a voluntary return, 21% ended with the child being taken to a non-Convention country and 8% with the child being taken to another Convention country.

57. 72% of the applications made under the Hague Convention had been finally dealt with by the time the data were collected, while only 38% of the European Convention applications had come to a conclusion.

58. 56% of the Hague applications and 47% of the European applications had been completed at the time of data collection.

59. It might also be noted that because of the relative ease with which the child's welfare can be raised as an issue under Art.10, the experience in England and Wales, at any rate, is that it is much more likely that a welfare report will be ordered in European Convention proceedings than in Hague Convention proceedings.

60. Only 3 of the completed “incoming” cases and 14 of the completed “outgoing” cases concerned access.

61. The average length of the “incoming” Anglo-German return applications was just under 6 weeks, while that of access applications was 35.5 weeks. See Lowe and Perry, op. cit. supra n.2.

62. Applications for return took between 1 and 46 weeks to complete, while completed applications for access lasted between 3 and 34 weeks.

63. It might also reflect the possibility that fathers experience greater difficulties than mothers in raising Art.13(b) defences that the child might be exposed to grave or psychological harm or otherwise be placed in an intolerable situation, if returned.

64. For further analysis of the operation of the Conventions between England and Wales and Germany see Lowe and Perry, op. cit. supra n.2.

65. 30% of the cases resulted in a judicial return or the registration of a custody order, while 13% resulted in the voluntary return of the children. 24% of the cases had not been resolved at the time of data collection, and a further 11% of the applications had been withdrawn by the applicant.

66. 89% of these cases involved a straightforward judicial return, 8% involved a judicial return by consent order and 3% involved registration of a custody order.

67. In 6 of these cases the court order was made by consent. The remaining 12% of cases were resolved as follows: the Unit refused 3% of the applications, 2% were voluntarily resolved without court intervention (but not necessarily by means of a voluntary return) and 1 case, or 0.6%, resulted in the registration of an access order. In 1 of these cases the applicant's instructions simply petered out, in 1 the file was closed because the child could not be traced, in 3% of the cases the child was taken to another Convention country and in 1 case the children had been taken to a non-Convention country. One case was dismissed and 1 stayed, but no further details were available, and finally, in I case (an access application from the US) the recorded result was that an existing access order had been forwarded to the US central authority. In 1 case, or 0.6%, the outcome was not known.

68. 72% of these were simple judicial returns, 20% were returns by consent order and 8% involved registration of a custody order.

69. In each case return of the eldest child was refused on the grounds of his or her objections to being returned.

70. In each case by consent. The requested central authority refused 4% of the cases, in 1% the file was dosed because the child could not be traced, 1% of the cases were closed after lack of instructions from the applicant, and an unspecified court order was made in less than 1% of the cases. In 1 case no action was taken.

71. See supra Section D.1 and Lowe and Perry, op. cit. supra n.2.

72. The CAU refused 5 cases, while among the “outgoing” cases, 7 applications were refused by central authorities.

73. Occurring in 43% of the “incoming” cases and 31% of the “outgoing” cases.

74. See also our analysis of the reasons for refusals to return in Anglo-German cases in Lowe and Perry, op. cit. supra n.2.

75. In 29% of “incoming” and 10% of “outgoing” cases.

76. In 14% of “incoming” and 10% of “outgoing” cases.

77. Each occurring in 10% of “outgoing” cases where return was refused.

78. The category “judicial returns” therefore comprises judicial returns, the registration of a custody or access order, the making of access orders, and those cases where the Convention proceedings were stayed, dismissed or withdrawn but a court order was nonetheless imposed upon the parties.

79. Including: judicial refusals under the Hague Convention, and refusals to register orders under the European Convention, but not those cases where a court order having the effect of a judicial refusal was made by consent.

80. The category “consensual resolutions” therefore comprises the following case outcomes: voluntary return, judicial return and judicial refusal by consent, the “other voluntary resolution” cases, and those cases where the Convention proceedings were stayed, dismissed or withdrawn but consent orders were made.

81. For present purposes, cases where the applicant ceases to give instructions have been classed as “withdrawn”.

82. Note that Fig.1 does not account for every case as some of the more unusual results recorded do not fit into any of these categories.

83. At supra Section D.4. Our follow up analysis has since revealed the following: with regard to incoming cases where the result was not known in 1% of cases, 48% were ordeed tobe returned, 8% were voluntarily returned, 8% ended in judicial refusals and 11% were withdrawn. With regard to outgoing cases the result was not known in 5% of cases, 23% were ordered to be returned, 20% were voluntarily returned, 9% ended in judicial refusals and 19% were withdrawn.

84. See Lowe and Perry, op. cit. supra n.2.

85. I.e. perhaps because of the time and costs involved. It has also been suggested to us that because, at any rate in England and Wales, it is so easy to make an application applicants may do so without really thinking, only to withdraw sometime later.

86. See Report, supra n.31, at paras.60 et seq. and Annexes I–III.

87. Ireland came a close second in terms of speed while, at the other end of the scale, Germany's system for dealing with Convention applications is particularly slow, see Loweand Perry, op. cii supra n.2.

88. Which at the time of writing amount to 21.

* The figures provided by the Australian Central Authority cover the period July 1995 to December 1996, i.e. 6 months more than those of the other countries.

† No information was provided on applications concerning children taken from Italy.

‡ The Portuguese figures do not include applications concerning children taken to or brought from France and Luxembourg. These applications are dealt with by means of bilateral agreements between Portugal and those countries.